Agentic security scanning (find/validate/fix vulns) on the Claude Code harness, reusing your Claude subscription. Fork of openai/codex-security. - View it on GitHub